The aim of this study is to investigate selected areas of the neurobiology of depression. In previous studies, we have observed that abnormality of noradrenergic and HPA axis dysfunction occur together in seriously depressed patients. We have pursued the study of the role of corticosteroids in depressive illness by examining the effect of exogenous steroid administration. We have found that orally administered dexamethasone produces selective effects on catecholamine function in depressed patients; in contrast to normal controls, depresed patients, particularly those with psychotic features, showed a significant dexamethasone-induced increase in plasma MHPG and a decrease in plasma HVA. These data suggesting abnormal corticosteroid-catecholamine interactions in depression are consistent with the possibility that hypercortisolemia itself may produce or enhance some of the biochemical changes and/or behavioral signs of depression. In a double-blind study of orally administered prednisone (80 mg/day x 5 days) to normal volunteers, we have further investigated steroid effects on the central nervous system. The relationship between stress, steroids, and mood, has been pursued in an experiment in which identical amounts of escapable and inescapable aversive noise stimuli are presented to subjects. Preliminary results suggest that inescapable but not escapable "stress" produces correlated mood and neuroendocrine response. This paradigm holds promise for examining not only stress response in normal subjects but also as a model of depression and as a tool to investigate biological diathesis in patients with current depressive illness and in subjects at risk for depression.